Tushar Dhyani
Research Scientist at Sony Europe
Stuttgart, Baden-Württemberg, Germany
Summary
👤
Senior🎓
Top SchoolTushar Dhyani is a Research Scientist at Sony Europe with 9 years of hands-on experience in AI, speech and audio ML, and production-ready model engineering. He progressed from internships and junior data science roles to research and applied work at Sony, including a master’s thesis that transitioned into a full-time research position. Tushar has practical expertise in audio data augmentation—contributing a Lambda augmentation to the widely used audiomentations library—plus experience in multilingual NLP and deploying conversational agents. He blends research rigor with product-minded engineering, having optimized model latency (e.g., a pruning technique that cut forward-pass time significantly) and automated data pipelines. Based in Stuttgart, he brings a pragmatic focus on making audio and dialogue models work robustly outside the lab.
9 years of coding experience
3 years of employment as a software developer
Bachelor's degree, Information Technology, Bachelor's degree, Information Technology at Madhav Institute of Technology and Science, Gwalior
English, Hindi, Gujarati
Stackoverflow
212reputation
11kreached
1answer
6questions
Github Skills (16)
machine-learning10
data-augmentation10
python10
numpy10
audio-processing9
deep-learning8
dspace8
dspic8
computer-vision6
elasticsearch6
data-mining6
feature-extraction6
data-cleaning6
feature-selection6
dataframe6
Programming languages (5)
Github contributions (5)
iver56/audiomentations
Jun 2022 - Oct 2022
A Python library for audio data augmentation. Useful for making audio ML models work well in the real world, not just in the lab.
Role in this project:
ML Engineer
ML EngineerContributions:1 review, 17 commits, 2 PRs in 3 months
Contributions summary:Tushar primarily contributed to the `audiomentations` Python library, which focuses on audio data augmentation for machine learning. They added a `Lambda` augmentation, allowing for custom operations on audio samples. Further commits involved modifying the `Lambda` function and integrating it into the library's documentation, and adding dependencies for better handling. The user also contributed by merging and adapting the code, and fixing documentation issues in demo files.
pythonalbumentationsaudio-datadspaudio-effects
Thanatoz-1/about
Dec 2017 - Dec 2024
Portfolio page
Contributions:10 PRs, 106 pushes, 9 branches in 7 years 2 months
cssjavascriptreactreactjs
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.